Can anybody stall Bitcoin for 72BTC per hour? ANSWER: NO

[phelix]

typical tx fee is 0.005 btc
https://en.bitcoin.it/wiki/Transaction_fees

~2400 txs fit into a block
http://bitcoin.stackexchange.com/questions/1747/bitcoin-block-size-what-are-the-rules

2400 * 0.0050001 * 6 blocks/hour  ~=   72

Miners will choose the tx with the highest fees. But to be really sure I just create 3000 tx with a fee of 0.011 every ten minutes. 33 BTC edit: 198 BTC

is that really all it takes?


Of course clients can increase fees. It would take time and user effort, though. I think with some clients it is not even possible for the user.


 

[1715057961]

1715057961

[1715057961]

1715057961

[1715057961]

1715057961

[1715057961]

1715057961

[grue]

fees get exponentially higher as the block fills up, so your fee assumption is wrong

[phelix]

fees get exponentially higher as the block fills up, so your fee assumption is wrong

hmm, could you please elaborate or give a link? you make it sound like there is some automatic system to do that but I never heard of that.

Would that be the case if I would do it right now?

[FreeMoney]

So someone spends 72/hr in order to make people pay .001 for tx? That sounds wonderful for miners and a tiny inconvenience for users. It's just ~1.5M dollars worth per month. Assuming you don't drive prices up at all.

[phelix]

So someone spends 72/hr in order to make people pay .001 for tx? That sounds wonderful for miners and a tiny inconvenience for users. It's just ~1.5M dollars worth per month. Assuming you don't drive prices up at all.

the establishment has plenty of dollars. also they can always print more


imagine bitcoin will stall for three days.

[FreeMoney]

So someone spends 72/hr in order to make people pay .001 for tx? That sounds wonderful for miners and a tiny inconvenience for users. It's just ~1.5M dollars worth per month. Assuming you don't drive prices up at all.

the establishment has plenty of dollars. also they can always print more


imagine bitcoin will stall for three days.

I suppose a short term inconvenience would not be too hard to pull off this way, but they would be supporting the bitcoin price and enriching bitcoin miners. An the public relations angle would be bad but not terrible, "Why am I not getting in a block?" Someone else is paying more than you.

Solutions exist and more will be made regarding setting your own fee and even replacing transactions to have a higher fee when they  don't get confirmed fast enough. So this is expensive, just an inconvenience, and going to become less effective over time even if the block size is not increased, which it probably will be.

[Bit_Happy]

Double F.U.D.
Can the network even handle a large volume of legit transactions?

[Bit_Happy]

Triple F.U.D.
Transactions take too long, Litecoins will become supreme?

[phelix]

So someone spends 72/hr in order to make people pay .001 for tx? That sounds wonderful for miners and a tiny inconvenience for users. It's just ~1.5M dollars worth per month. Assuming you don't drive prices up at all.

the establishment has plenty of dollars. also they can always print more


imagine bitcoin will stall for three days.

I suppose a short term inconvenience would not be too hard to pull off this way, but they would be supporting the bitcoin price and enriching bitcoin miners. An the public relations angle would be bad but not terrible, "Why am I not getting in a block?" Someone else is paying more than you.

Solutions exist and more will be made regarding setting your own fee and even replacing transactions to have a higher fee when they  don't get confirmed fast enough. So this is expensive, just an inconvenience, and going to become less effective over time even if the block size is not increased, which it probably will be.

media would eat us alive like vultures. imagine all tx would go to an address that is known to be connected with highly illegal activity.

mobile clients would basically stop working.

[phelix]

ebay's market cap is ~ $87189480000 PayPal ~ 1/3  =  $29063160000 = mc

198btc  *  24h * 365days * 35$/btc  =   60706800    ~=  mc * 0.2%

if I got it right paypal could stall bitcoin for a year at 0.2% of their market cap.

even worse, as price would fall stalling would become cheaper. guessed cost to run her into the ground: $100,000

As bitcoin price would drop they could even profit from it (going short).


I always hate when I realize something disturbing about Bitcoin.

[btcusr]

typical tx fee is 0.005 btc
https://en.bitcoin.it/wiki/Transaction_fees

~2400 txs fit into a block
http://bitcoin.stackexchange.com/questions/1747/bitcoin-block-size-what-are-the-rules

2400 * 0.0050001 * 6 blocks/hour  ~=   72

Miners will choose the tx with the highest fees. But to be really sure I just create 3000 tx with a fee of 0.011 every ten minutes. 33 BTC edit: 198 BTC

is that really all it takes?


Of course clients can increase fees. It would take time and user effort, though. I think with some clients it is not even possible for the user.


 

That's so evil. Such adjustments could encourage mining but not now.  

[DeathAndTaxes]

There is no such thing as stall.  People would pay higher fees and life would go on.  This tactic however would massively enrich miners and cause a boom to the exchange rate. 

[goatpig]

ebay's market cap is ~ $87189480000 PayPal ~ 1/3  =  $29063160000 = mc

198btc  *  24h * 365days * 35$/btc  =   60706800    ~=  mc * 0.2%

if I got it right paypal could stall bitcoin for a year at 0.2% of their market cap.

even worse, as price would fall stalling would become cheaper. guessed cost to run her into the ground: $100,000

As bitcoin price would drop they could even profit from it (going short).


I always hate when I realize something disturbing about Bitcoin.

You don't understand the mechanics of Bitcoin. Your assumption is that tx fees will remain static. Bitcoin is built to function without the coinbase reward, only on tx fees. Competition for tx fees is natural and already part of the system. This isn't an attack, it's just using Bitcoin as intented. The more transactions are processed, the more expensive spots in blocks get. This "attack" would only increase the BTC/$ exchange rate at an exponential rate.

Simple example: Assume that your attacker emits 2400 tx per block with an 0.005 fee. It is fair to say this would increase the minimum fee by 0.001 BTC, as anyone who wants to get his tx through would just pay 0.006. Now assume the attacker is a lunatic and is going to fill his next tx batch with the new min fee. He would essentially increase the fee by 0.001 BTC per block, on every block. A week down the road, the fee would be about 0.15 BTC and this guy would need about 50k BTC a day just to fill the blocks.

This practice would have 2 effect: First, he would the biggest consumer (read: supporter) of BTC out there. After a couple weeks of this stupidity, the "attacker" would represent 90% of daily BTC trade volume, pumping the market cap up 10 folds.

Second, I doubt anyone with a meaningful BTC tx would be afraid of even a 0.5 BTC fee. You were talking about Paypal. Their fee is like 5% per transaction. That would be a tx as small as 10BTC with a 0.5BTC fee. However, a 0.5BTC fee implies the attacker needs 170k BTC a day to pull his little trick. With such a big buyer, we'll break $500/BTC easily and it would cost him short of $100mil to pull his move on a daily basis.

[laughingbear]

I am not worried about anyone buying 72 btc every hour on the hour, and then giving it away to miners, hell... I encourage it!

[misterbigg]

I guess if you're willing to spend close to $2 million USD a month to "stall" Bitcoin then yeah it would work but I would love to be a miner in that scenario.

[btcusr]

yes, that's about 1,734,480 BTC a year.
@ 50 USD exchange rate per BTC, it would be $86,724,000 a year

[qwk]

I think most of the posters in this thread are missing the point, really.

Of course, the bitcoin protocol itself is in no way endangered.
And of course, the default client and probably most full-blown implementations of alternative clients can easily cope with said "attack".

But if somebody with the money and the motivation really pulled this of for, say, a week, here's what i believe would happen:

- people with only tiny amounts of bitcoins would not be able to experiment with it, because they could not afford the fees.
- people with implementations of a client that does not let them adjust the transaction fee would be unable to spend their coins.
- people with a high volume of transactions of small size would face a huge increase in costs.
- Satoshi Dice ... hmmm, i wonder ... but actually, i don't care.

After a few hours, it'd be field day for haters on slashdot etc.
A few days, the media.
Weeks, the bitcoin "community".

The miners might have fun for a day or two, but then?


All in all, once the dust settles, it'd not be the end of bitcoin.
But it might very well be a setback of at least the same impact as the MtGox hack.


So, what's the conclusion? Check each and every implementation of bitcoin clients, be they online wallets, mobile apps, lightweight, heavyweight, whatever applications for one thing:

Is the user able to adjust the transaction fee? If not, fix this.
This should be "best common practice" for all developers of bitcoin clients.

[GoldenWings91]

Such an 'attack' would make the price of BTC skyrocket. In order to sustain this 'attack' a lot of bitcoins would need to be purchased. The laws of supply and demand dictates if there is a limited supply and a large demand price will go up. At 72BTC needed per hour the demand for BTC would be enormous. The longer this 'attack' lasts the more bitcoins are in demand the more expensive they become. This would make such an attack unsustainable for even the richest entities. Even though transactions won't be able to process for a while I doubt anybody would complain when seeing the value of their bitcoins skyrocketing.

[Gavin Andresen]

The default block-filling algorithm that most miners are running is:

+ Fill up part of the block with the highest transactions, regardless of fees
+ Then fill up the rest of the block with as many fee-paying transactions as possible, highest fee-per-kilobyte first.

... so flooding the network with high-fee transactions won't "stall Bitcoin."  Well, except for people playing SatoshiDice or doing something else that results in lots of low-priority fee-paying transactions (and even there, they could always opt to pay a little more in transaction fees).

[phelix]

The default block-filling algorithm that most miners are running is:

+ Fill up part of the block with the highest transactions, regardless of fees
+ Then fill up the rest of the block with as many fee-paying transactions as possible, highest fee-per-kilobyte first.

... so flooding the network with high-fee transactions won't "stall Bitcoin."  Well, except for people playing SatoshiDice or doing something else that results in lots of low-priority fee-paying transactions (and even there, they could always opt to pay a little more in transaction fees).

I guess you mean "highest priority transactions".

Glad to hear that. It makes the attack mostly pointless. Hope miners will keep up that policy.


I am always happy when something I thought could be a problem for Bitcoin turns out not to. 


Still, as qwk suggests, it would be good to be able to modify fees in every client. A notification to the user that the payment was delayed with a question if the fee should be raised might be nice.

[FreeMoney]

Such an 'attack' would make the price of BTC skyrocket. In order to sustain this 'attack' a lot of bitcoins would need to be purchased. The laws of supply and demand dictates if there is a limited supply and a large demand price will go up. At 72BTC needed per hour the demand for BTC would be enormous. The longer this 'attack' lasts the more bitcoins are in demand the more expensive they become. This would make such an attack unsustainable for even the richest entities. Even though transactions won't be able to process for a while I doubt anybody would complain when seeing the value of their bitcoins skyrocketing.

Yes, and if the attack was actually working they wouldn't be able to buy more since no one could get coins to the exchanges, rofl.

[phelix]

Such an 'attack' would make the price of BTC skyrocket. In order to sustain this 'attack' a lot of bitcoins would need to be purchased. The laws of supply and demand dictates if there is a limited supply and a large demand price will go up. At 72BTC needed per hour the demand for BTC would be enormous. The longer this 'attack' lasts the more bitcoins are in demand the more expensive they become. This would make such an attack unsustainable for even the richest entities. Even though transactions won't be able to process for a while I doubt anybody would complain when seeing the value of their bitcoins skyrocketing.

Yes, and if the attack was actually working they wouldn't be able to buy more since no one could get coins to the exchanges, rofl.

besides of what gavin said: do you seriously think a couple of days of bitcoin not working would not affect the price?

[dooglus]

+ Fill up part of the block with the highest transactions, regardless of fees

I guess you mean "highest priority transactions".

It would appear so.  From src/main.cpp:

Code:

    // How much of the block should be dedicated to high-priority transactions,
    // included regardless of the fees they pay
    unsigned int nBlockPrioritySize = GetArg("-blockprioritysize", 27000);

So by default only the first 27k of each block is dedicated to the highest priority transactions.

[qwk]

besides of what gavin said: do you seriously think a couple of days of bitcoin not working would not affect the price?

Maybe calling it "not working" is just a little too harsh.

A couple of days where the average user experiences massive delays for at least a part of his/her transactions and where bitcoin based services face unexpected extra costs for their infrastructure, is what i would call it.

And that seems tough enough.

[phelix]

+ Fill up part of the block with the highest transactions, regardless of fees

I guess you mean "highest priority transactions".

It would appear so.  From src/main.cpp:

Code:

    // How much of the block should be dedicated to high-priority transactions,
    // included regardless of the fees they pay
    unsigned int nBlockPrioritySize = GetArg("-blockprioritysize", 27000);

So by default only the first 27k of each block is dedicated to the highest priority transactions.

hmmm....  roughly 65 transactions....   edit: currently about 14.4%

besides of what gavin said: do you seriously think a couple of days of bitcoin not working would not affect the price?

Maybe calling it "not working" is just a little too harsh.

A couple of days where the average user experiences massive delays for at least a part of his/her transactions and where bitcoin based services face unexpected extra costs for their infrastructure, is what i would call it.

And that seems tough enough.

agreed

[solex]

+ Fill up part of the block with the highest transactions, regardless of fees

I guess you mean "highest priority transactions".

It would appear so.  From src/main.cpp:

Code:

    // How much of the block should be dedicated to high-priority transactions,
    // included regardless of the fees they pay
    unsigned int nBlockPrioritySize = GetArg("-blockprioritysize", 27000);

So by default only the first 27k of each block is dedicated to the highest priority transactions.

hmmm....  roughly 65 transactions....   edit: currently about 14.4%

besides of what gavin said: do you seriously think a couple of days of bitcoin not working would not affect the price?

Maybe calling it "not working" is just a little too harsh.

A couple of days where the average user experiences massive delays for at least a part of his/her transactions and where bitcoin based services face unexpected extra costs for their infrastructure, is what i would call it.

And that seems tough enough.

agreed

So, is this yet another plus for an algorithmic block size limit, which could then absorb such an "attack" and still allow most other legitimate transactions through in a timely manner?

[AbsoluteZero]

You cannot stall Bitcoin.

But For 2400 satoshis Per Block In transaction fees, you can make It that only paying
(or High priority) transactions make It into the Block.

This would cost about 0.12 cents of FIAT Per day

[gmaxwell]

But For 2400 satoshis Per Block In transaction fees, you can make It that only paying
(or High priority) transactions make It into the Block.

Surprise: The people working on this stuff are not actually idiots. Very tiny fees are treated as zero, precisely to close off that sort of funny business.

[AbsoluteZero]

But For 2400 satoshis Per Block In transaction fees, you can make It that only paying
(or High priority) transactions make It into the Block.

Surprise: The people working on this stuff are not actually idiots. Very tiny fees are treated as zero, precisely to close off that sort of funny business.

Ok, how about 2400 transactions per block from old coins?

[Digigami]

How does one go about changing or modifying the rules that the client they run follows? Are there options in the config file to do this or does this need to be done on a source code level and then compiled by the user?

I'm curious how out of reach it is for average joe - somewhat knowledgeable individuals this is, or is this exclusive to the experts who can re-write the source?

I run a fairly well connected Satoshi client node. I used to solo mine from it, quite a long time ago but have been a pooler since. I opted to keep the node on line to help support the network, as it is run at very little cost to me, I don't use the wallet attatched to it, so really it's only purpose is to help relay transactions, and of course help distrubute the chain to re syncing/new clients as I am able to maintain a relatively large number of connections. On and off I will see my node in blockchain's hub-nodes list, and frequently it will be listed on transactions as the relaying IP.

Basically, I am curious to know what options a node operator such as myself has at my disposal should I decide I want my node to limit relaying or broadcasting low priority tx's such as satoshidice activity, or other low priority transactions? If I was still mining solo or if I was a pool operator, and still used the Satoshi client, to change the rules of which transactions I would include in blocks once again is this adjusted via user friendly flags or options or does one need to be a coder to change the rules?

I don't think I even want to modify my client, at least not this well connected node because basically I feel it is my single greatest contribution to Bitcoin as a whole, not because I wouldn't want to do more, but because my expertise is limited, and other then being a faithful user and advocate I am otherwise not the type who can add much to what's going on.

In the future, if the developers as a whole have not introduced changes which address the emerging issue with spam transactions, I would almost feel it my duty to all Bitcoiners that I adjusted my node to at least reduce it's participation in the problem, and ignore/not relay transactions which are troublesome. I doubt it will come to this, I have faith in our dev team that the problem with be adequately addressed before it comes down to users such as myself taking action, but I would like to know it is at least possible to do something should the need arise.

Appreciate any feedback you all have on the topic. Thanks

[poly]

Such an 'attack' would make the price of BTC skyrocket. In order to sustain this 'attack' a lot of bitcoins would need to be purchased. The laws of supply and demand dictates if there is a limited supply and a large demand price will go up. At 72BTC needed per hour the demand for BTC would be enormous. The longer this 'attack' lasts the more bitcoins are in demand the more expensive they become. This would make such an attack unsustainable for even the richest entities. Even though transactions won't be able to process for a while I doubt anybody would complain when seeing the value of their bitcoins skyrocketing.

Absolutely. Default clients sending the general fee won't go through, so there would be less bitcoins sold on the market. This would continually increase the price of BTC and such an attack is unsustainable.

However, I'll leave a conspiracy theory of the current rally is caused by the supposed establishment buying up bitcoins to halt it.

[btcusr]

The default block-filling algorithm that most miners are running is:

+ Fill up part of the block with the highest transactions, regardless of fees
+ Then fill up the rest of the block with as many fee-paying transactions as possible, highest fee-per-kilobyte first.

Point 1 can also be manipulated. Anyways, I am just moving BTC between my addresses. For others it would look like spending, but in reality I own all 6K addresses in a block. so, it is like, for about 3k transactions, we would need 3000 BTC to prevent transactions below 1 BTC to enter into blockchain. plus, whatever we are paying for transaction fee.

We don't need 3000 BTC for every block, it is just one time amount required, and we are just moving this amount around.

[Sukrim]

I guess Gavin meant "highest priority transactions"? As coin age is factored in there, you can pull this off just once.

[phelix]

I guess Gavin meant "highest priority transactions"? As coin age is factored in there, you can pull this off just once.

he definitely meant priority. see doglus' post above.

[Gavin Andresen]

Yes, I definitely meant priority. Highest priority transactions (transferring lots of old coins) get included in blocks first under the default block-filling rules.

And also notice that I said "most miners are..." There are at least a few big mining pools that have their own idiosyncratic ways of deciding which transactions get into blocks, including private deals with big exchanges/merchants/etc.

Also note that because finding blocks is a random process the Bitcoin network "stalls" for an hour every three weeks or so, with no blocks found.

My guess is that if an attacker tried to monopolize block space most of us wouldn't even notice. If you're really worried about it, then encourage some big mining pool(s) to have a completely different block-filling strategy ("randomly select from the memory pool" would be easy to implement).

[phelix]

Yes, I definitely meant priority. Highest priority transactions (transferring lots of old coins) get included in blocks first under the default block-filling rules.

And also notice that I said "most miners are..." There are at least a few big mining pools that have their own idiosyncratic ways of deciding which transactions get into blocks, including private deals with big exchanges/merchants/etc.

Also note that because finding blocks is a random process the Bitcoin network "stalls" for an hour every three weeks or so, with no blocks found.

My guess is that if an attacker tried to monopolize block space most of us wouldn't even notice. If you're really worried about it, then encourage some big mining pool(s) to have a completely different block-filling strategy ("randomly select from the memory pool" would be easy to implement).

An hour is acceptable but not a day even if it is only 85%.

But again you have a good argument. After a couple of hours pools can react and change the block-filling strategy. They make less short-term but hopefully enough are wise enough to realize it will pay of long term.

Maybe an emergency block-filling strategy should be implemented and tested so that pools can switch quickly should there be need.

[phelix]

[...]
The old settings were default, with the 250,000 byte limit, 27k for high-priority (regardless of fee).  The new settings we're trying (subject to change, and not on all servers yet) is 500kB block max size, no reserved space for no-fee transactions with high priority, and a minimum size set to 50 kB to grab high priority/no-fee transactions if there aren't that many unconfirmed paid transactions.

so much for that...

[phelix]

With the block size limit currently in fact being 256kb... this is even cheaper.

A chunk of old coins and 50btc per hour. Somebody will give this a try for a day or a couple of days sooner or later.

Hopefully we will get a higher block size limit soon so it gets a little more expensive.

[phelix]

psst. tx fee assumption in op is off by a factor of 10. don't tell PP.

[phelix]

Looks like somebody finally gives it a try: https://bitcoinmagazine.com/21125/bitcoin-network-still-backlogged-tens-thousands-unconfirmed-transactions-causing-delays/

[johnyj]

Could this attack itself partially initiated by large mining pools? Because they get part of the spam coin back through fee, it does not cost them too much, but it will make lots of coins on market occupied during the attack, thus raise the exchange rate. With more and more coins trapped in unconfirmed status, the coin supply on market will become less and less, a great help for the speculators who are pumping up the price at the same time

[Cryddit]

I note that this "stress test" coincides with a rise in price and an increase in "buy" volumes on major exchanges.  

Amid all the digital dust caused by unconfirmed transactions, someone (or more likely someseveral) is buying massive amounts of bitcoin.

I wonder if the current "attack" is being done by (or on behalf of) big buyers.  It would be intended to occupy the news and speculation while a move that would otherwise set off a more major price spike gets executed.  Pure conspiracy theory, of course.

Note:  As I write this my memory pool shows 74,581 unconfirmed transactions, which is 181 megabytes of backlog.

[Cryddit]

Alternatively this could be someone trying to prevent a (VERY LARGE) zero-conf transaction from getting into the block chain.   By design that attack shouldn't work, because a VERY LARGE transaction would inevitably rise in priority during a single day until it were included in the "freebie" section.  That is, in fact, why Satoshi put a freebie section in the block composition code in the first place.

But now that we have a bunch of miners ignoring the freebie section, sellers are at least theoretically vulnerable to it.  Which, essentially, means the advice about always waiting for at least a couple confirmations is essential.

Cryddit

[phelix]

I had always thought of someone with a big put option on Bitcoin but these would work, too:

Could this attack itself partially initiated by large mining pools? Because they get part of the spam coin back through fee, it does not cost them too much, but it will make lots of coins on market occupied during the attack, thus raise the exchange rate. With more and more coins trapped in unconfirmed status, the coin supply on market will become less and less, a great help for the speculators who are pumping up the price at the same time

The risk for a panic seems pretty high. Otherwise it could simply be large pools trying to create a proper fee market.

I note that this "stress test" coincides with a rise in price and an increase in "buy" volumes on major exchanges. 

Yeah, a whale trying to buy low. But it would make more sense to wait with buying until the problem surfaces, maybe even sell a bunch to create a panic.